Solid oxide fuel cell (SOFC) technology has considerable promise for meeting many long-term, world-wide energy requirements. This technology uses cheap, readily available fuels, such as hydrogen, methane or simple alcohols. Its highly efficient power production capabilities give rise to outstanding opportunities for this technology as a future energy source for many commercial, aerospace and defense-related applications. In order to be successful, however, methods must be developed for fabricating solid oxide fuel cells reliably and cost effectively.
One fuel cell design employs thin ceramic components in a honeycomb structure. The honeycomb structure consists of thin layers of four components: 1) anode; 2) electrolyte; 3) cathode; and, 4) interconnect. Although this design results in markedly improved performance, it also multiplies the technical difficulties in fabricating these multicomponent substrates. The usual method of fabrication is to form thin layers of the pure components and to heat the multilayer structure at a high temperature.
Of particular importance is the ceramic material that is used as an interconnect material. Four materials that are currently used in solid oxide fuel cell designs include Mg doped LaCrO.sub.3, Sr doped LaMnO.sub.3, Ni/ZrO.sub.2 and CaO/ZrO.sub.2. Of these materials, Mg doped LaCrO.sub.3 presents significant difficulties due to its low availability, high cost and a high sintering temperature. Typical sintering conditions required to sinter LaCrO.sub.3 or Mg doped LaCrO.sub.3 to full density (or closed porosity) are extremely low oxygen partial pressures and a temperature of 1750.degree. C. The low oxygen partial pressure is needed to reduce volatilization of chromium due to oxidation which has been found to inhibit the sintering of this material. See L. Groupp and H. U. Anderson, J. Amer. Ceram. Soc., Vol. 59, No. 9-10, 449-50 (1976). The high sintering temperature is very restrictive since heating the multilayer structure to temperatures necessary to sinter the LaCrO.sub.3 or Mg doped LaCrO.sub.3 often results in failure (cracking) of the multilayer substrate. The development of lanthanum chromite powders capable of sintering to full density, i.e., 95% of theoretical density, below 1700.degree. C. is critical to reduce Cr volatilization problems and improve fuel cell fabrication.
One example of a method of manufacturing lanthanum chromite (LaCrO.sub.3) electrodes is disclosed in U.S. Pat. No. 3,974,108. This patent discloses that strontium doped LaCrO.sub.3 can be produced by preparing a slurry of lanthanum oxide, strontium carbonate and chromic acid, drying said slurry in air and then preferably calcining at a temperature of 1200.degree. to 1500.degree. C. to give a strontium doped LaCrO.sub.3 powder. Sintering of this material occurs at temperatures above 1700.degree. C.
Attempts to lower the sintering temperature of this material by adding fluxes have had limited success. A key drawback of this approach is the deleterious effect the added flux(es) can have on the materials present in the other layers, especially on their electrochemical properties and thermal expansion coefficients.
An alternative approach is to use sol-gel technology to prepare high surface area, reactive LaCrO.sub.3 powders that sinter to full density below 1700.degree. C. The reduction in sintering temperature is accomplished by controlling the size, composition, morphology, homogeneity, and reactivity of the material. Such control is achieved by tailoring the solution chemistry and powder processing parameters. One such method for preparing LaCrO.sub.3 (lanthanum chromite) precursors is disclosed in U.S. Pat. No. 3,330,697. This process involves dissolving two or more metal salts (i.e., carbonates, hydroxides) in citric acid and ethylene glycol. The resulting sol is then filtered, dried to a gel, and calcined to remove the organics. However, this process results in some residual carbon being present in the material which can have a detrimental effect on the sintering properties of the material.
An alternate procedure has been disclosed by C.N.R. Rao et al. "Synthesis of Complex Metal Oxides Using Hydroxide, Cyanide and Nitrate Solid Solution Precursors", Journal of Solid State Chemistry, Vol. 58, 29-37 (1985). The method consists of coprecipitating a solid solution of isostructural La and Cr hydroxides and produces highly homogeneous intimate mixtures of these two elements. The solid solution precursor, LaCr(OH).sub.6, can be converted to LaCrO.sub.3 by subsequent calcination at 850.degree. C. for 12 hours.
Specifically, Rao et al. teach the precipitation of La.sub.0.5 Cr.sub.0.5 (OH).sub.3 by adding an aqueous nitrate solution of the metal ions to a sodium hydroxide (NaOH) solution. Rao also teaches that in order to reduce sodium contamination, the gel must be washed extensively with hot water. As will be shown in greater detail herein, we attempted to prepare lanthanum chromite by Rao's method and found that even after extensive washing with hot water, a small quantity (about 0.02 weight percent) of sodium was present in the gel. While seemingly low, even this amount of sodium is sufficient to markedly change the solid state chemistry of the gel. Further, if a magnesium doped material is desired, the extensive washing required to remove sodium ions can also result in marked leaching of magnesium from the gel resulting in gels with poorly controlled stoichiometry.
Rao et al. additionally disclose that ammonium hydroxide can be used to precipitate the hydroxides. However, only La.sub.0.5 Al.sub.0.5 (OH).sub.3 was prepared using ammonium hydroxide and Rao states that when divalent metals are desired, the method using ammonium hydroxide does not work. Thus, Rao teaches that ammonium hydroxide cannot be used to precipitate a compound containing divalent metal such as magnesium.
In marked contrast to Rao's teachings, we have surprisingly found that ammonium hydroxide can be used to precipitate an hydroxide salt having the empirical formula LaCr.sub.x A.sub.1-x (OH).sub.6, where A is a divalent metal such as magnesium, strontium or barium and which can be converted to LaCr.sub.x A.sub.1-x O.sub.3. We have thus prepared a compound by a process which the prior art taught would not produce such a compound.
In addition, the process of the instant invention produces a product which has several advantages over the product of the prior art. For example, we have found that the process of the prior art gave a gel which upon calcining at 600.degree. C. produced a multiphase mixture of La.sub.2 CrO.sub.6, Cr.sub.2 O.sub.3 and La.sub.2 O.sub.3. Surprisingly, the process of the instant invention gave a single phase product, e.g. LaCrO.sub.4 or LaCr.sub.0.95 Mg.sub.0.05 O.sub.4 .multidot.0.04H.sub.2 O. More surprisingly, the product of the instant invention has a huttonite structure. A huttonite structure is an unusual structure first described by P. Hutton in "Huttonites, A New Monoclinic Thorium Silicate", American Mineralogy, Volume 36, 60-65 (1951). Although it is known that LaCrO.sub.4 has a huttonite structure, (H. Schwarz, Z. Anorg. Allgem. Chem. 322, 1-14 (1963); M. D. Vasileha et al. Dapov. Akad. Nauk Ukr. RSR, Ser. B: Geol., Khim. Biol. Nauki 1977, (5), 410-13.) the prior art does not disclose that LaCr.sub.0.95 Mg.sub.0.05 O.sub.4 .multidot.0.04H.sub.2 O has a huttonite structure.
Upon further calcining the above products at 900.degree. C. in air, the product of the prior art produced a powder having a La.sub.2 CrO.sub.6 and a LaCrO.sub.3 phase, whereas the product of the instant invention produced only LaCrO.sub.3. The physical characteristics of these powders were also observed to be drastically different. The process of the prior art gave a powder with a melted type morphology, and 5-10 micron particles which could not be reduced by intensive milling. In marked contrast, the process of the instant invention gave a powder with a porous and vermicular microstructure and 5-10 micron particles which were quickly reduced to particles of about 0.5 microns.
The smaller particle size and single phase of the powder gives three very desirable results. First the smaller particle size of the powder allows the powder to be sintered into a ceramic article (having a density of at least 95% of theoretical density) at a lower temperature than a powder with larger particles. Secondly, a multiphase powder, i.e., La.sub.2 CrO.sub.6 and LaCrO.sub.3, will affect the thermal expansion coefficient of the final ceramic articles. If the thermal expansion coefficient does not match that of the other components of the fuel cell the entire fuel cell may fail upon repeated thermal cycling. Therefore, it is critical to have a single phase powder. Third, the multiphase powder does not have the proper electrical properties for use as an interconnect material, while the single phase powder does possess the desired electrical properties.
Thus, the instant invention differs surprisingly and significantly from the prior art in that the instant invention: 1) provides a process using ammonium hydroxide for preparing a ceramic powder precursor having a huttonite structure; 2) provides a huttonite compound which can be used to prepare a ceramic powder which has a single phase and smaller particles than that of the prior art and 3) said ceramic powder is sinterable to 95% of theoretical density at a lower temperature than that of the prior art.